Journal of Gender Studies
ISSN: (Print) (Online) Journal homepage: www.tandfonline.com/journals/cjgs20
Spotify (Un)wrapped: how ordinary users critically
reflect on Spotifys datafication of the self within
creative workshops
Taylor Annabell & Nina Vindum Rasmussen
To cite this article: Taylor Annabell & Nina Vindum Rasmussen (26 Nov 2024): Spotify
(Un)wrapped: how ordinary users critically reflect on Spotify’s datafication of the self within
creative workshops, Journal of Gender Studies, DOI: 10.1080/09589236.2024.2433674
To link to this article: https://doi.org/10.1080/09589236.2024.2433674
© 2024 The Author(s). Published by Informa
UK Limited, trading as Taylor & Francis
Group.
Published online: 26 Nov 2024.
Submit your article to this journal
Article views: 729
View related articles
View Crossmark data
Full Terms & Conditions of access and use can be found at
https://www.tandfonline.com/action/journalInformation?journalCode=cjgs20
RESEARCH ARTICLE
Spotify (Un)wrapped: how ordinary users critically reect on
Spotify’s datacation of the self within creative workshops
Taylor Annabell
a
and Nina Vindum Rasmussen
b
a
Faculty of Law, Economics and Governance, Utrecht University, The Netherlands;
b
London School of Economics
and Political Science, London, UK
ABSTRACT
Each year, Spotify nudges users to share aesthetically pleasing data
stories ‘wrapped’ and repackaged from their listening behavior. This
article approaches Spotify Wrapped as an annual algorithmic event,
dened as a moment in time in which there is a collective orientation
towards a particular algorithmic system and its associated data. It oers
a methodological contribution to research on datacation of music taste
and identities through the development of a workshop format aimed at
ordinary Spotify users. The workshop delivers insights into practices of
datacation and the normative assumptions baked into Spotify data
stories. Drawing on a data feminist framework, we outline three inter-
connected but distinct creative exercises, which take participants on an
analytical journey. We combine feminist arts-based research methodol-
ogies with critical reection and the walkthrough method to centre
people’s experiences and equip them to analyze dierent layers of
Wrapped. Our theoretical and methodological approach seeks to desta-
bilize the logics of data extraction that further Spotify’s commercial aims
and its associated claims of ‘knowing us’ through the aggregation of user
data. As such, our workshop transforms the marketing campaign into
a site for critical reection on Wrapped as an algorithmic event.
ARTICLE HISTORY
Received 24 January 2024
Accepted 19 November 2024
KEYWORDS
Algorithmic culture;
algorithmic events; creative
methods; datafication;
Spotify
Introduction
The end of November marks Spotify Wrapped season: the time of year when Spotify users confront
a digest of their streaming habits from January to November. When users open the app in late
November, they are invited to ‘play’ their Wrapped breakdown, presented through short, consecu-
tive videos similar to the ‘stories’ format used by other platforms such as Instagram. The in-app
experience is available for about one month, but a bespoke playlist with one’s top 100 songs remains
available as a time capsule. Although the format of these data stories changes in each iteration, the
logic and purpose persist: inviting users to participate in an annual ‘algorithmic event’ by sharing
personalized listening stats and comparing them to broader trends. We dene an algorithmic event
as a moment in time in which there is a collective orientation towards a particular algorithmic system
and its associated data. Yet what the platform’s claims of knowing ‘you’ mean to ordinary people is
yet to be empirically explored. In this article, we outline a workshop method that enables research-
ers, educators and students to examine how ordinary users engage with algorithmic events like
Wrapped.
CONTACT Taylor Annabell t.annabell@uu.nl Faculty of Law, Economics and Governance, Utrecht University, Achter Sint
Pieter 200, 3512 HT Utrecht, The Netherlands
JOURNAL OF GENDER STUDIES
https://doi.org/10.1080/09589236.2024.2433674
© 2024 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.
0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. The terms on which
this article has been published allow the posting of the Accepted Manuscript in a repository by the author(s) or with their consent.
Spotify’s wrap-ups began in 2013 under the name ‘Year in Review’, a webpage that unveiled
individualized listening data to users (Alagiah, 2022). In 2016, it was rebranded as ‘Spotify Wrapped’.
The Wrapped email was opened by 30 million users, drove over 1 million streams on Spotify and
featured in 1,548 pieces of media coverage. By 2022, Spotify displayed in their Q4 2022 update that
the Wrapped campaign engaged 150 million monthly active users and grew 30% from 2021 across
111 markets (Spotify, 2023). Part of this popularity stems from the way Spotify adds new personalized
and interactive features to these data stories every year. In 2021 users were presented with their
‘Audio Aura’, which captured the two top ‘moods’ and visualized them as a swirling gradient of color
(Figure 1).
In 2022, Spotify replaced this with two new curated data stories: ‘Your Listening Personality’ and
‘Your Audio Day’ (Figure 2). That year also put motion design at the centre of the campaign. Upon
clicking on the main Wrapped feature, vibrantly colored monograms in dierent shapes would
overlap and interlock to help make the presentation of the stats more dynamic. The designers of the
campaign explained how motion design gave each layer within a monogram a distinct sense of
personality, behavior and ‘motion language’ which was supposed to signal how ‘we’re all unique’
(Alagiah, 2022).
The social aspect is a core component of Wrapped as an algorithmic event and, therefore, key to
our proposed method. The personalized data stories are designed to be shared and discussed on
social media: Each card in the Wrapped story has a ‘Share this story’ button to facilitate this process,
turning the phenomenon into an organic marketing engine for Spotify. The platform amplies the
signicance of its aggregated metrics in outdoor ads, where global listening patterns surface
through games like mazes, word searches, connect the dots and ll-in-the-blanks. By playfully
connecting shared listening patterns to lifestyle, Spotify naturalizes the logic of extracting user
data to make claims about society and identity. This strategy serves the economic project of
surveillance capitalism, which ‘unilaterally claims human experience as free raw material for transla-
tion into behavioral data’ (Zubo, 2019, p. 8). Central to our method for critically engaging and
analysing Wrapped is providing space for participants to expose and interrogate their relationship to
these logics.
Figure 1. Screenshots of the Spotify Wrapped in-app experience in 2021 (Spotify, 2021).
2 T. ANNABELL AND N. V. RASMUSSEN
Our proposed method enables workshop facilitators (who may be researchers or teachers) and
participants to collaboratively interrogate Wrapped as an algorithmic event. We argue that a critical
and arts-based approach signicantly advances our understanding of the way algorithmic logics and
social dynamics intersect during algorithmic events. Importantly, the workshop aords ordinary
users the opportunity to reect on how they experience and feel about data capture as well as
platforms’ interpretation of such data. We have developed this method as part of our ongoing
research project, ‘Spotify (Un)wrapped’, which is guided by the following questions: How do users
see their behavior as shaped by the collective orientation towards Wrapped as an algorithmic event?
How do users perceive the assumptions Spotify appears to be making about their identity, taste and
lifestyle? Finally, can users resist this kind of datacation of listening habits, and if so, how?
This article does not provide a thematic analysis of participant contributions in response to these
questions. Instead, we detail our methodological approach and demonstrate how it generates rich
responses to such questions. In particular, we utilize arts-based methods to cast new light on the
datacation of music taste as well as ongoing discussions in critical algorithm studies of how to know
‘the algorithm’. The next sections outline our theoretical and methodological frameworks. As we
detail later in this article, some of the terms introduced here play an important role as conceptual
tools in the workshop itself.
Datacation of music taste, identity and lifestyle
We situate our project within existing research that investigates the implications of turning cultural
production and consumption into data; a process referred to as ‘datacation’ (Mayer-Schönberger &
Cukier, 2013). By investing heavily in machine learning capabilities over the years, Spotify has sought
to deliver an increasingly personalized listening experience for users (see Stål, 2021). Spotify does so
by processing the ongoing ow of user data and making predictive assessments about behavioral
patterns and trends. As music listening is turned into data, companies like Spotify have, therefore,
gained the capacity to make claims about ‘knowing us’ in new and intimate ways (Webster, 2023,
p. 2141).
Figure 2. Screenshots of the Spotify Wrapped in-app experience in 2022 (Spotify, 2022).
JOURNAL OF GENDER STUDIES 3
Personal playlists form a cornerstone of this operation. Spotify’s playlist feature enables users to
act as content curators of their music consumption and actively manipulate and maintain these
collections over time (Hagen, 2015, p. 644). Playlists also provide Spotify with a detailed under-
standing of the semantic meaning of particular songs and how they relate to specic situations and
aective states (e.g. ‘Songs I like to vibe to while gardening’ and ‘Relaxing music for anxiety and
panic attacks’, which are titles of two real playlists on Spotify). As such, Spotify users perform data
labelling work by grouping songs together in personal playlists. By utilizing a machine learning
technique called collaborative ltering, Spotify employs this information to recommend titles to
subscribers with similar tastes and listening behavior (Stål, 2021). Put dierently, your taste is not
your own but something that, to a degree, is shared with countless others (Seaver, 2022, p. 25).
Morris and Powers (2015) point out that streaming services like Spotify sell ‘branded musical
experiences that target certain styles of musical access, discovery and use’ (p. 109). They argue that
such services try to reect people’s attitudes, everyday habits and feelings towards music back to
them. In many ways, Spotify wants to be the musical backdrop in people’s everyday lives. That leads
Pedersen (2020) to argue that Spotify pushes users towards a form of ‘ubiquitous listening’, which
favors music’s ‘functional’ value (i.e. listening to music in the background while doing other things),
‘rather than music’s aesthetic value and the depth of the emotions it produces as an object of
contemplation and attentive listening’ (p. 87; see also Hesmondhalgh, 2022 for a critical overview of
this debate).
Following this view, Eriksson et al. (2019) suggest that Spotify aims to reorganize the consump-
tion of music ‘around behaviors, feelings, and moods, which are channelled through curated playlists
and motivational messages that change several times a day’ (p. 5). Under the headline '
', Spotify
prompted one of the authors of this paper to click on a playlist titled ‘Sad Girl Summer’. The playlist
thumbnail featured a pink ice lolly melting against a rosy background, accompanied by the
subheading ‘It’s okay to feel sad in the sunshine’. Later that day, the interface advised that same
user to switch it up with a playlist called ‘Serotonin’, promising ‘100% good vibes’. This quick glance
at the interface reveals how Spotify makes inferences about the gender, daily activities, uctuating
moods and aesthetic preferences of users. As such, Spotify can be said to play a role in ‘programming
our most mundane and intimate everyday activities’ (Burgess et al., 2022, p. 50). Our research
examines how users respond to this everyday programming as part of the algorithmic event
Spotify Wrapped.
Although Wrapped is the subject of media commentary, how people experience and encounter
this algorithmic event has received limited attention in academic literature. An exception to this is
the analysis of #spotifywrapped tweets carried out by Burgess et al. (2022). They refer to the
Wrapped phenomenon as a ‘Spotify data sele’, which implies there is a tension between looking
inward and looking outward. However, Burgess et al. emphasize the ‘active, knowing and ambivalent
relationship users have to Spotify’s data-driven story about who they are as music consumers’ (p. 54).
This complexity shines through in users’ reections on the ‘glitchy relationships between actual
everyday practices and the algorithm’s logics’, such as nursery rhymes or white noise tracks making it
to the top tracks list. For Burgess et al., tweets on these glitches expose how users make sense of the
discrepancy between algorithmic proles and the complexities of everyday life. Posting these
reections via the public hashtag #spotifywrapped furthermore allows users to take part in ‘collec-
tive intimacy’ and ‘processes of collective learning and debate about algorithmic culture’ (p. 55). Our
proposed method enables people to unwrap the dierent layers of datacation in a critical and
engaging way. To this end, we build on existing research on everyday experiences and interpreta-
tions of algorithmic systems.
Making sense of algorithms in everyday life
The repackaging of user data is integral to Wrapped and can be situated within the
conguration of the self through data assemblages. A data assemblage is a socio-technical
4 T. ANNABELL AND N. V. RASMUSSEN
system composed of several apparatuses and elements that entwine, develop, and mutate
over time and space (Kitchin, 2014, pp. 24–25). As Lupton (2016) puts it, personal data
assemblages, which are responsive to new inputs and interpretations, represent a dynamic
selfhood that is ‘distributed between dierent and changing datasets’ (p. 115). In a similar
vein, Cheney-Lippold (2011) refers to ‘algorithmic identities’ produced by algorithms that
infer categories of identity based on data. Such categorization of identity is projected onto
individuals outside of their control (p. 176). Even so, some users attempt to control the
conguration of their algorithmic identity by manipulating ‘the ways in which their personal
data are collected, archived and used’ (Lupton, 2016, p. 117). Our workshop takes an interest
in such practices, along with the ways users understand the role played by algorithms in
subject formation.
We also draw on Prey’s (2018) notion of ‘algorithmic individuation’, which he denes as ‘a
dynamic socio-technical process engaged in enacting the individual’ (p. 1095). Prey notes the
prevalence of algorithmic individuation on platforms like Spotify and how algorithms play an
increasingly important role in subject formation: ‘The “others” we interact with are increasingly
algorithms reecting back categorized images of our self’ (p. 1096). Additionally, Prey explains what
sets algorithmic individuation apart from earlier forms of individuation is the black-boxed nature of
specic recommendations (pp. 1096–1097). Despite this knowledge asymmetry, we argue that
Spotify actively promotes algorithmic individuation by pushing users to look inward through data
analysis, especially during Wrapped season.
Wrapped also sits within a broader context of algorithmically curated memories. Work in digital
memory studies has explored how platform ‘Memories’ or ‘Year in Review’ mobilize models of
memory through the algorithmic ‘remembering’ of previously shared digital traces (see Jacobsen
& Beer, 2021; Prey & Smit, 2019). Jacobsen (2022) calls attention to the way that algorithms impose
narrative structures onto images. By grouping and resurfacing selected images as ‘Memories’ for
users, the feature attributes meaning to the past and implies it ‘knows’ what is meaningful to the
user, which is akin to the Wrapped rhetoric. By drawing on interviews and focus groups, Jacobsen
(2022) shows how such memories are ‘sociotechnically produced and felt in everyday life’ (p. 2872).
For Bucher (2017), aective encounters with algorithms as part of everyday life involve what she
refers to as ‘the algorithmic imaginary’. This concept is central to our workshop method as it
emphasizes how people perceive and navigate their everyday experiences with algorithms. This
sense-making process often plays out socially, which researchers have called attention to with terms
like ‘folk theories’ (e.g. Eslami et al., 2016; Siles et al., 2020) and ‘algorithmic gossip’ (Bishop, 2019).
Following this line of thought, Gandini et al. (2023) argue that algorithmic systems construct users as
digital and data subjects, but human users also subjectivize the algorithms in return. Users engage in
this ‘counter-subjectivation’ in response to the individuation enacted by algorithmic systems: ‘In so
doing, users engage in a process of “othering” of the algorithm(s), in an eort to dierentiate their
agency from that of “the algorithm”’ (p. 418). Gandini et al. emphasize focus groups as a useful
avenue for studying user interactions with this ‘algorithmic other’ (pp. 421–422). We agree and
propose that feminist and creative approaches can push this method even further.
Feminist and arts-based approaches to digital culture
Our creative workshops are guided by the seven principles of data feminism outlined by
D’Ignazio and Klein (2020, 2024): examine power, challenge power, rethink binaries and hierar-
chies, elevate emotion and embodiment, consider context, embrace pluralism and make labor
visible. Intersectional feminism takes an analytical interest in the root causes of structural
inequalities and how power imbalances can be challenged, rebalanced and changed. To this
end, their data feminist framework examines and challenges the way the power of data is
wielded unequally. That includes challenging the gender binary, ‘along with other systems of
counting and classication that perpetuate oppression’ (D’Ignazio & Klein, 2024, p. 8). Data
JOURNAL OF GENDER STUDIES 5
feminism also elevates and synthesizes multiple forms of knowing. This means recognizing the
knowledge that arises from people as ‘living, feeling bodies in the world’ and centralizing ‘local,
Indigenous, and experiential ways of knowing’ (pp. 9–10). Following the work within critical data
studies more broadly, data feminism furthermore situates data within its social context. It high-
lights the labor of data science and recognizes that ‘all data is shaped by unequal social
relations’ (p. 12).
These principles underpin the design of the Spotify (Un)wrapped workshop. The principles of
‘examining power’ and ‘rethinking binaries’ are built into all of the creative exercises, which we
demonstrate later in this article. The workshop is also orientated towards capturing ‘pluralism’,
‘context’ and the ‘emotion and embodiment’ of participants in how they approach Wrapped as an
algorithmic event. As such, we contribute to the data feminist framework by detailing how it can be
applied in a creative workshop setting to examine algorithmic events.
We draw inspiration from arts-based methods to evoke taken-for-granted knowledge of algo-
rithmic systems, data assemblages and algorithmic events. In line with Lupton and Watson (2021),
we argue that creative workshops can ‘elicit the aective and multisensory contexts of people’s
feelings, practices and imaginaries concerning their digital data’ (p. 463). This approach promotes
a more embodied form of knowledge making which has traditionally been neglected in academic
contexts. As Schwittay (2021) argues, ‘feminist scholars have shown that this subordination is often
connected to the devaluation of feminine forms of knowing and being in the world’ (p. 42). Arts-
based methods are therefore intimately linked with feminist research. For instance, such methods
have been used in feminist research with girls and young women as a way to explore – often through
crafting experiences and feelings related to advertising, relationships, sexual violence and harass-
ment (see Renold, 2018; Ringrose et al., 2021; Venema & Lobinger, 2017).
In their ‘Algorithmic Autobiographies and Fictions’ workshops, Bishop and Kant (2023) ask
participants to respond to their ‘algorithmic selves’ generated through social media advertising
data. After locating lists of ad topics, participants are invited to draw an algorithmic self and write
about meeting this algorithmic self. Bishop and Kant note how this often sees participants dealing
with inconsistencies of data and complexities of their layered identities. By engaging with their data
and target proles, participants are prompted to ‘question how social media platforms are selling
their identities and interests to marketers’ (p. 1026). Our proposed workshop takes a similar interest
in people’s creative interpretations of their algorithmic selves.
Siles et al. (2020) also use creative drawing within their empirical work examining recommenda-
tions on Spotify. They utilize the concept of folk theories to analyze how participants produce ‘rich
pictures’ that explain how they perceive the workings of algorithmic recommendations on Spotify.
Bringing together these images with discussions in focus groups and interviews, Siles et al. demon-
strate how participants mobilize dierent folk theories, which are used as resources to frame the
agency of the user. Our project extends the interest of Siles et al. in understanding how people
theorize and creatively engage with Spotify’s data practices. We are especially interested in the
algorithmic imaginaries, algorithmic identities and folk theories that emerge when Spotify users
engage with Wrapped in a creative setting. The next section outlines how our interactive workshop
format integrates feminist and creative approaches to facilitate such insights.
Running the Spotify (Un)wrapped workshop
Our two-hour workshop consists of ve main elements: (1) introducing Wrapped and our conceptual
toolkit, (2) decoding the Wrapped listening characters, (3) analysing Wrapped or the Spotify interface
using a modied version of the walkthrough method (Light et al., 2018), (4) assembling a ‘Spotify
data sele’ (Burgess et al., 2022) in the form of a collage, and (5) a nal wrap-up where participants
reect on their analytical journey. Participants collaborate in groups of ve and engage in plenary
discussions following each element.
6 T. ANNABELL AND N. V. RASMUSSEN
The creative workshop is designed for ordinary people with an interest in Wrapped and familiarity
with Spotify or similar music platforms. Although we draw on theories and methods used in
academic research, the workshop does not necessitate prior knowledge or experience. So far, we
have hosted nine workshops with more than 200 participants in the United Kingdom from 2023 to
2024. Eight workshops were with undergraduate and postgraduate students from the humanities
and social sciences. The nal workshop was part of a public engagement event, attracting a more
general audience.
As researchers, we approach data collection during the workshop in keeping with
Markham’s (2006, p. 7) proposition that ‘all methods decisions are in actuality ethics deci-
sions’. In every workshop, we obtain informed consent from each participant to audio record
their contributions in the plenary discussions, which are anonymized in research outputs. We
also photograph and publish materials they produce, oering them the option to take these
home should they wish. Although participants are invited to bring screenshots of their
Wrapped, we do not ask participants to send us these documents as a form of data.
Instead, we focus on their interpretations of their own data, prioritizing their analytical
insights.
Positioning ordinary users as co-analysts
The Spotify (Un)wrapped workshop seeks to avoid reproducing the patterns of data extraction from
users that platforms perpetuate. Instead, it is committed to the feminist value of reciprocity and
beneting the communities engaged with (Johnston & MacDougall, 2021, p. 3). Reciprocity involves
the concerted eorts of researchers to ‘give back’ to participants, for instance in the form of time,
resources or both (Ellingson, 2017, p. 49). First, we see the workshop as a place for participants to
collaboratively gain tools to navigate and potentially resist everyday datacation. When funding
permits, we also oer participants refreshments and a live music performance to enrich their
experience and acknowledge their contributions. Finally, we seek to ‘collaborate with participants
as equals, speak with rather than for participants’ and ‘develop solutions to problems identied by
participants’ (Ellingson, 2017, p. 49).
Accordingly, we position our participants as co-analysts following the work of Robards and
Lincoln (2017) and Markham (2021). Fundamental to Robards and Lincoln’s development of the
‘scroll back method’ is the way participants narrate their digital traces as they ‘scroll back’ through
a specic platform. This method recognizes how participants’ sense-making and narration is an act of
analysis, which generates interview data that can be further analyzed by researchers. Markham
(2021) oers a similar sentiment with her call for research that enables people to ‘act more as
researchers of their own lived experience’ (p. 400). Such a proposition builds on the valuable
empirical work on how ordinary people perceive and experience algorithmic systems and dataca-
tion by recognizing their contributions as analytical. In the following sections, we highlight how this
is operationalized within our workshop design and reect on the insights made possible by these
exercises.
Introducing Wrapped and our conceptual toolkit
The workshop opens with an introductory presentation to contextualize our approach to Wrapped.
We present some background to its history and the most recent iteration. We also outline the
concepts of algorithmic identities (Cheney-Lippold, 2011), algorithmic imaginaries (Bucher, 2017)
and algorithmic gossip (Bishop, 2019). Sketching our theoretical apparatus establishes a shared
vocabulary for the remainder of the workshop. In addition, it is in keeping with how we position
participants’ contributions as part of academic enquiries into everyday encounters with algorithmic
systems.
JOURNAL OF GENDER STUDIES 7
We then move to an introductions round. In their groups, each participant shares one reection
on Wrapped or another instance of repackaged music data. Each group presents a common theme
that emerged from their discussion to the wider group. Across our workshops, this introductory
round has generated responses that range from celebration of Spotify’s data capture and interpreta-
tion to critical and resistant readings. As one participant expressed: ‘We really look forward to Spotify
Wrapped because it gives us a little insight into what we’ve listened to in the year or like our
experiences throughout the year’. This initial discussion often establishes core themes and topics
that animate the rest of the workshop.
Decoding the listening characters of Spotify Wrapped
The rst creative exercise invites participants to analyze the Wrapped listening characters. In
2022, Spotify developed ‘Listening Personality’ types to portray how you listen to music,
independent of what music you like’ (McDonald, 2023). Echoing the Myers-Briggs MBTI
Personality Types, four sets of binary characteristics (Familiarity/Exploration, Timelessness/
Newness, Loyalty/Variety, Commonality/Uniqueness) are used to form 16 ‘Listening
Personalities’. For example, the combination of the attributes Exploration, Timelessness,
Loyalty and Uniqueness are labelled ‘The Maverick’ personality, which is further described
as ‘You know who you are as a listener. While everyone’s bathing in the mainstream you’re
frolicking in that sidestream’. In 2023, Spotify called this feature ‘Me in 2023’ and assigned
users one of 12 characters such as the ‘Roboticist’, ‘Vampire’ and ‘Fanatic’. Like the 2022
iteration, ‘Me in 2023’ makes claims about users’ behavior, taste and identity.
We approach these listening characters as a repackaged version of algorithmic identities
(Cheney-Lippold, 2011), as they oer a glimpse into the categories of identity Spotify infers
upon its users. The exercise encourages participants to critically discuss this categorization and
theorize about their ‘algorithmic other’ (Gandini et al., 2023). It also invites reection on whether
platforms can meaningfully distil data points into simple characters, given that data assemblages
are ‘always mutable, dynamic, responsive to new inputs and interpretations’ (Lupton, 2016,
p. 115).
Each group receives a deck of the most recent listening character cards. This gives participants an
alternative view of the categorization, allowing them to assess the range of characters and their
descriptions. This is unlike the in-app experience of Wrapped where the user can only access their
one assigned character. Participants are encouraged to make notes as a way to document their analysis.
We provide the following questions to spark discussion: What connections does Spotify establish
between music listening and taste? What do these labels suggest about how data is being collected?
Reecting on your Wrapped, how do you feel about this categorization? By centring their personal
experience, this nal question prompts participants to consider the process of algorithmic individua-
tion propelled by Wrapped (Prey, 2018).
Across our workshops, we have observed how this exercise draws attention to the static and
reductive nature of these listening characters. By restricting the user to one personality and collating
user data, some of our participants experience Wrapped as collapsing diachronic, dynamic listening
behavior, which they consider core to how their listening takes place. Here, the glitchiness of
Wrapped (Burgess et al., 2022) goes beyond a disconnect in Spotify’s representation of listening
behavior from their lived experience to also critique the processes of categorization that rely on
shallow assessments of ‘true’ listening behavior. For instance, one of our groups noted that the
categorizations are ‘very generic and simplistic (16 labels how do they collect the data?)’, followed
by the question ‘is this accurate?’ (Figure 3). For others, this exercise raises questions of gendered
algorithmic identities, evidenced through the language and imagery used within the cards. Finally,
this exercise shows participants how their collective folk theories (Eslami et al., 2016; Siles et al., 2020)
can generate useful insights into this process of categorization despite its opaque nature.
8 T. ANNABELL AND N. V. RASMUSSEN
Walking through Wrapped
The second creative exercise is a modied version of the walkthrough method developed by Light
et al. (2018). This method involves mapping the aordances of an app by systematically moving
through the navigation and ow of the interface. Within the technical walkthrough, the researcher
assumes the position of the user and methodically observes characteristics such as the user interface
arrangement, functions and features, textual content and tone and symbolic representation. In
tracing how the app congures relations between actors, the researcher is encouraged to be
sensitive towards ‘how the app constructs conceptions of gender, ethnicity, ability, sexuality and
class’ (Light et al., 2018, p. 891). The walkthrough method has also been developed as an exercise for
teaching environments to compare two apps that serve similar purposes (see Duguay & Gold-Apel,
2023).
In the Spotify (Un)wrapped workshop, the comparative app walkthrough takes a slightly
dierent form. Instead of contrasting dierent apps, participants are invited to compare the
interface of two dierent users. Our version remains committed to the orientation to Spotify’s
materiality as well as perceptions of aordances and built-in inequalities, but it is carried out
within the time restraints of the workshop. Participants work in pairs to walk through either the
Spotify app or their screenshots or recordings of the Wrapped in-app experience, focusing on the
way data, personalization and identity are signalled and constructed. Most participants will only
have access to screenshots of their Wrapped. For that reason, we provide an audiovisual
recording of our own Wrapped data stories that participants can analyze if they wish. We see
this as another way to atten the power relationship between us as researchers and our
participants. Many participants tend to engage with this recording to remind themselves of
the structure, features and overall feel of the in-app Wrapped experience, even if the specic
artists and stats dier from their own.
Figure 3. Participants analyze the 2022 ‘listening personalities’ in February 2023.
JOURNAL OF GENDER STUDIES 9
As part of the analytical process, they produce a visual representation of their observations about
the ways data and identity entangle (Figures 4 and 5). This usually takes the form of a diagram,
owchart or drawing. These visual representations provide an alternative way of documenting the
process of engaging in a technical walkthrough and nudge participants to consider the ow of
Figure 4. Walkthrough diagram of the 2022 Wrapped experience from February 2023 workshop.
Figure 5. Walkthrough diagram of the 2022 Wrapped experience from October 2023 workshop.
10 T. ANNABELL AND N. V. RASMUSSEN
navigation alongside the visibility of content. By directly contrasting the navigation through the
Spotify app or the Wrapped screenshots and recordings, participants observe dierences and simila-
rities in the interface, which they use to establish how Spotify acts as a mediator of personalization.
Among other things, this exercise enables participants to identify dierences in color schemes and
types of recommendations, for instance in relation to underlying conceptions of gender.
In our experience, participants tend to connect details made visible through the walkthrough
with insights from earlier discussions. For instance, Figure 4 proposes that the ‘Listening Personality’
remediates the 16 personalities test and, in doing so, our need for a ‘sense of belonging’. Meanwhile,
Figure 5 uses dierent colors and quotation marks to distinguish between descriptions of elements
with their own interpretations. These two diagrams demonstrate how our modied walkthrough can
be used to analyze complex socio-technical systems like Wrapped within a short timeframe. We see
this inclusion of a visual element as a contribution to the further development and application of the
walkthrough method.
Creative responses to Wrapped as an algorithmic event
Drawing on the feminist and arts-based approaches to digital culture outlined earlier, the nal
exercise sees participants producing a physical artefact. This creative exercise aords participants
agency to express their experiences and thoughts in ways that dier from other exercises.
Participants construct a Wrapped collage or ‘Spotify data sele’ (Burgess et al., 2022) using material
objects such as CDs, vinyl records, magazines, glitter and markers. We emphasize to participants that
this can be a response, critique or a rearticulation of their Wrapped. In other words, the exercise
requires participants to produce an artefact that connects to the algorithmic event, but is not limited
to Spotify’s data practices.
Prior to the workshop, participants complete a form to conrm their attendance and tell us their
top Wrapped song. While participants assemble their collages, we stream a playlist of these songs
curated for that workshop and, when funding permits, a musician may perform a short set. The
creative exercise also adheres to D’Ignazio and Klein’s (2020) principle of ‘elevating emotion and
embodiment’. Importantly, participants do not produce the artefacts to discursively unpack them
afterwards. Following Vacchelli (2018), we argue that an embodied approach should resist the use of
collage as an elicitation strategy: ‘The artefact itself (the collage), the narrative used to discuss it, the
memories and the emotions evoked by the research participants are used to “hear the stories” the
participants tell through their bodies’ (p. 174).
Through their collaging, participants convey a variety of feelings, experiences and perceptions.
One creative artefact (Figure 6) reimagines the ‘top artist’ metric by imbuing it with personal,
subjective experiences. The collage overlays dierent geographies and temporalities (as represented
through the clock) to indicate the meaningfulness of listening to music. It thereby hints at the
limitations of the Wrapped articulation of ‘knowing the self’. Another example is Figure 7, which uses
smashed CD pieces to represent fragmentary shards of identity that the platform reects back to the
individual. In a playful way, this artefact challenges the premise of Wrapped and questions whether
the platform can use data to make meaningful claims about identity. Figure 7 critically deconstructs
Wrapped, while Figure 6 visualizes the experience of listening over time in a way that transcends
Spotify’s data stories.
From our experience, several participants produce artefacts that resonate more closely with the
format and visual identity of the Wrapped data stories. Others may centre on a celebration of
individual music tastes constructing a fan-identity. We propose that the openness and exibility of
the creative exercise lend itself to such diversity: It aords opportunities for critical reection on the
datacation and personalization that underpins Spotify, but it is also a space for participants to
express their relationship to music listening more broadly. This includes the functional, aesthetic and
emotional value of music.
JOURNAL OF GENDER STUDIES 11
Figure 6. Creative artefact produced at one of the Sheffield workshops in 2023.
Figure 7. Creative artefact produced at one of the Sheffield workshops in 2023.
12 T. ANNABELL AND N. V. RASMUSSEN
Analytical journey in the Spotify (Un)wrapped workshops
The workshop adheres to what Markham (2013) calls ongoing playful engagement with remix
methods. We envisage the structure of the workshop three distinct exercises that build on each
other – as an analytical journey for participants to investigate dierent aspects of Wrapped. The data
generated in the form of audio recordings of plenary discussions and participant creations (dia-
grams, artefacts, notes) can be analyzed by researchers in multiple ways. It can be approached
through insights generated by separate exercises, but can also be brought together thematically,
making connections across the dierent exercises.
The production of creative artefacts in the nal exercise functions as the culmination of
the participants’ analytical journey throughout the workshop. The previous exercises become
part of the material that participants choose to draw upon in their creative responses. For
instance, one collaborative piece between two participants (Figure 8) reects on their
exploration during the walkthrough of how categories of recommendation intersected with
gender. Similarly, when discussing Figure 9, another participant refers to the rst exercise,
noting: ‘Spotify with the personality things is trying to make everyone feel like “oh,
they’re special in their music taste”, but we are all listening to Taylor Swift and crying about
our lives’. The artefact also highlights the power of the recommender system by replacing
‘here’s what you liked’ with ‘here’s what Spotify suggested’, indicating that Spotify acts as
a co-constructor and mediator of taste. Finally, the participant hints at the celebratory
aspects of understanding the self through behavioral data with the creation of a ‘musical year
in a ipbook’. The multiplicity of these narratives speaks to the complicated feelings towards
Wrapped as participants navigate the algorithmic intervention into their experience of music.
Figure 8. Collaborative artwork made by two participants at a London workshop in 2023.
JOURNAL OF GENDER STUDIES 13
In keeping with our emphasis on participants as co-analysts, we invite participants to reect on the
design of the workshop. On a slide, we include key terms like algorithmic events and pose possible
prompt questions: ‘Spotify “knowing us” can we resist? If so, how?’ and ‘What questions still need to
be asked, and what needs to be explored?’ Participants write their answers on sticky notes, which we
use to shape future iterations of the workshop and our research agenda.
Discussion and conclusion
Every year, Spotify holds a data mirror to its users with the release of Spotify Wrapped. We regard
Wrapped as an algorithmic event that naturalizes the process of extracting data and repackaging it
to make claims about identity and taste. Our workshop method allows participants to explore how
this process is injected with assumed value for users, and how that maps onto the platform’s eorts
to attract and retain subscribers. It provides dierent entry points for creatively and critically
unpacking datacation and Spotify’s claims about ‘knowing us’ through our data (Webster, 2023).
Building on feminist arts-based and creative workshop research methodologies (e.g. Bishop &
Kant, 2023; Lupton & Watson, 2021), this article has introduced the Spotify (Un)wrapped workshop.
Our proposed method consists of reective and creative exercises that allow participants to act as co-
analysts. We propose that our interconnected but distinct exercises take participants on an analytical
journey that culminates in the production of a creative artefact. This workshop setup provides space
for people to encounter their ‘Spotied’ selves from dierent angles and critically explore what they
think of such data-driven interactions.
Our application of a data feminist framework (D’Ignazio & Klein, 2020, 2024) contributes to
existing scholarship on the datacation of music consumption (e.g. Burgess et al., 2022; Eriksson
Figure 9. Creative artefact produced at a London workshop in 2023.
14 T. ANNABELL AND N. V. RASMUSSEN
et al., 2019; Morris & Powers, 2015; Pedersen, 2020; Seaver, 2022). More specically, a data feminist
approach allows us to centre the experiences, feelings and perspectives of participants in research,
which subverts the Wrapped premise of constructing identity and music taste for users based solely
on their listening data.
Participants and workshop facilitators collaboratively examine the power dynamics between
Spotify and its users. That includes a focus on how conceptions of identity (e.g. with regard to
gender, ethnicity, ability, sexuality and class) can be observed in Wrapped. We propose that the
focus on normative assumptions baked into Wrapped in particular has the potential to generate
critical insights that go beyond existing approaches. As we have shown throughout the article,
the workshop format enables researchers to generate a wealth of empirical data. However, it
brings with it the limitation of a deeper understanding of individual participants’ experiences.
The workshop could be combined with other methods such as interviews or focus groups with
fewer participants.
Through our workshops, we seek to transform algorithmic events like Wrapped into an
opportunity for critical reection on the underlying algorithmic logics, power structures and
social dynamics. Our concept of the ‘algorithmic event’ and application of the data feminist
framework to the creative workshop setting could be taken up by researchers to examine other
entanglements of algorithmic systems and daily life. Our reections on the workshop and
exercises gesture towards the complicated feelings towards Wrapped and how Spotify shapes
music listening and identity construction. The workshop makes space for participants to grapple
with the claims that Wrapped data stories can and do reect the ‘self’ in a meaningful way. As we
have argued throughout this article, the insights generated with this methodological approach
both utilize and advance existing theorizations of algorithmic identities (Cheney-Lippold, 2011),
algorithmic individuation (Prey, 2018), personal data assemblages (Kitchin, 2014; Lupton, 2016)
and the ‘Spotify data sele’ (Burgess et al., 2022). Analysing the nuanced responses and creative
artefacts produced during such workshops oers another avenue for examining how people
applaud, interrogate and resist the collective orientation towards algorithmic systems during
algorithmic events.
Acknowledgments
Both authors have equally contributed to this article; authorship is in alphabetical order. This project was rst presented
at ‘Algorithms for Her?’ in 2023, and we wish to thank the organising committee for hosting such a supportive, critical
conference that generated invaluable feedback from participants. We also want to thank the special issue editors and
reviewers whose thoughtful questions and insights have been instrumental in rening our work.
Disclosure statement
No potential conict of interest was reported by the author(s).
Funding
The author(s) disclosed receipt of the following nancial support for the research, authorship, and/or publication of this
article: This work was supported by the Arts and Humanities Research Council [grant number AH/R012679/1], LSE100
and the LSE Festival.
Notes on contributors
Taylor Annabell is a Postdoctoral Researcher in the ERC Starting Grant HUMANads project at Utrecht University. Her
research interests include platform governance and inuencer cultures, everyday experiences of datacation and
gendering of digital memory work.
JOURNAL OF GENDER STUDIES 15
Nina Vindum Rasmussen is an LSE100 Fellow at the London School of Economics and Political Science. Her research
investigates how digital and algorithmic technologies impact the way we work, create and live, especially in the context
of media production and consumption.
ORCID
Taylor Annabell http://orcid.org/0000-0002-6384-6754
Nina Vindum Rasmussen
http://orcid.org/0009-0004-1107-167X
Data availability statement
The participants of this study did not give written consent for their data to be shared publicly. Due to the ethical and
condential restrictions of the research, supporting data is therefore not available.
Departmental ethics approver
Dr Jillian Terry, London School of Economics and Political Science, Ref: 262076.
Research Ethics Facilitator on behalf of the Arts and Humanities Research Ethics Panel: Dr Alexander Miller Tate,
King’s College London, Ref: LRS/DP-22/23-34,727.
References
Alagiah, M. (2022, November 30). How Spotify’s Wrapped campaign for 2022 came together. It’s nice that. https://www.
itsnicethat.com/features/spotify-wrapped-campaign-identity-2022-graphic-design-301122
Bishop, S. (2019). Managing visibility on YouTube through algorithmic gossip. New Media & Society, 21(11–12),
2589–2606. https://doi.org/10.1177/1461444819854731
Bishop, S., & Kant, T. (2023). Algorithmic autobiographies and ctions: A digital method. Sociological Review, 71(5),
1012–1036. https://doi.org/10.1177/00380261221146403
Bucher, T. (2017). The algorithmic imaginary: Exploring the ordinary aects of Facebook algorithms. Information
Communication & Society, 20(1), 30–44. https://doi.org/10.1080/1369118X.2016.1154086
Burgess, J., Albury, K., McCosker, A., & Wilken, R. (2022). Everyday data cultures. Polity Press.
Cheney-Lippold, J. (2011). A new algorithmic identity: Soft biopolitics and the modulation of control. Theory, Culture &
Society, 28(6), 164–181. https://doi.org/10.1177/0263276411424420
D’Ignazio, C., & Klein, L. (2020). Data feminism. MIT Press.
D’Ignazio, C., & Klein, L. (2024). Data feminism for AI. Proceedings of the 2024 ACM Conference on Fairness, Accountability,
and Transparency (FAccT ’24) (pp. 100–112). USA: Association for Computing Machinery. https://doi.org/10.1145/
3630106.3658543
Duguay, S., & Gold-Apel, H. (2023). Stumbling blocks and alternative paths: Reconsidering the walkthrough method for
analyzing apps. Social Media + Society, 9(1). https://doi.org/10.1177/20563051231158822
Ellingson, L. L. (2017). Embodiment in qualitative research. Routledge.
Eriksson, M., Fleischer, R., Johansson, A., Snickars, P., & Vonderau, P. (2019). Spotify teardown: Inside the black box of
streaming music. MIT Press.
Eslami, M., Karahalios, K., & Sandvig, C. (2016). First I ‘like’ it, then I hide it: Folk theories of social feeds. Proceedings of the
20 CHI conference on human factors in computing systems (CHI ‘16) (pp. 2371–2382), USA. http://dx.doi.org/10.1145/
2858036.2858494
Gandini, A., Gerosa, A., Giurè, L., & Keeling, S. (2023). Subjectivity and algorithmic imaginaries: The algorithmic other.
Subjectivity, 30(4), 417–434. https://doi.org/10.1057/s41286-023-00171-w
Hagen, A. N. (2015). The playlist experience: Personal playlists in music streaming services. Popular Music and Society, 38
(5), 625–645. https://doi.org/10.1080/03007766.2015.1021174
Hesmondhalgh, D. (2022). Streaming’s eects on music culture: Old anxieties and new simplications. Cultural Sociology,
16(1), 3–24. https://doi.org/10.1177/17499755211019974
Jacobsen, B. (2022). When is the right time to remember? Social media memories, temporality and the kairologic. New
Media & Society, 26(5), 2872–2888. https://doi.org/10.1177/14614448221096768
Jacobsen, B., & Beer, D. (2021). Social media and the automatic production of memory. Bristol University Press.
Johnston, K., & MacDougall, C. (2021). Enacting feminist methodologies in research toward reproductive justice.
International Journal of Qualitative Methods, 20, 1–10. https://doi.org/10.1177/16094069211016157
Kitchin, R. (2014). The data revolution: Big data, open data, data infrastructures and their consequences. Sage.
16 T. ANNABELL AND N. V. RASMUSSEN
Light, B., Burgess, J., & Duguay, S. (2018). The walkthrough method: An approach to the study of apps. New Media &
Society, 20(3), 881–900. https://doi.org/10.1177/1461444816675438
Lupton, D. (2016). The diverse domains of quantied selves: Self-tracking modes and dataveillance. Economy and
Society, 45(1), 101–122. https://doi.org/10.1080/03085147.2016.1143726
Lupton, D., & Watson, A. (2021). Towards more-than-human digital data studies: Developing research-creation methods.
Qualitative Research, 21(4), 463–480. https://doi.org/10.1177/1468794120939235
Markham, A. (2006). Ethic as method, method as ethic: A case for reexivity in qualitative ICT research. Journal of
Information Ethics, 15(2), 37–54. https://doi.org/10.3172/JIE.15.2.37
Markham, A. (2013). Remix culture, remix methods: Reframing qualitative inquiry for social media contexts. In N. Denzin
& M. Giardina (Eds.), Global dimensions of qualitative inquiry (pp. 63–81). Left Coast Press.
Markham, A. (2021). The limits of the imaginary: Challenges to intervening in future speculations of memory, data, and
algorithms. New Media & Society, 23(2), 382–405. https://doi.org/10.1177/1461444820929322
Mayer-Schönberger, V., & Cukier, K. (2013). Big data: A revolution that will transform how we live, work, and think. John
Murray.
McDonald, G. (2023, January 5). What’s a “listening personality”? Spotify R&D. https://engineering.atspotify.com/2023/
01/whats-a-listening-personality
Morris, J., & Powers, D. (2015). Control, curation and musical experience in streaming music services. Creative Industries
Journal, 8(2), 106–122. https://doi.org/10.1080/17510694.2015.1090222
Pedersen, R. (2020). Datacation and the push for ubiquitous listening in music streaming. MedieKultur: Journal of Media
and Communication Research, 36(69), 71–89. https://doi.org/10.7146/mediekultur.v36i69.121216
Prey, R. (2018). Nothing personal: Algorithmic individuation on music streaming platforms. Media Culture & Society, 40
(7), 1086–1100. https://doi.org/10.1177/0163443717745147
Prey, R., & Smit, R. (2019). From personal to personalized memory: Social media as mnemotechnology. In Z. Papacharissi
(Ed.), A networked self and birth, life, death (pp. 209–224). Routledge.
Renold, E. (2018). ‘Feel what I feel’: Making da(r)ta with teen girls for creative activisms on how sexual violence matters.
Journal of Gender Studies, 27(1), 37–55. https://doi.org/10.1080/09589236.2017.1296352
Ringrose, J., Regehr, K., & Zarabadi, S. (2021). Feminist craftivist collaging: Re-mattering the bad aects of advertising. In
D. Lupton & D. Leahy (Eds.) Creative approaches to health education (pp. 105–122). Routledge.
Robards, B., & Lincoln, S. (2017). Uncovering longitudinal life narratives: Scrolling back on Facebook. Qualitative
Research, 17(6), 715–730. https://doi.org/10.1177/1468794117700707
Schwittay, A. (2021). Creative universities: Reimagining education for global challenges and alternative futures. Bristol
University Press.
Seaver, N. (2022). Computing taste: Algorithms and the makers of music recommendation. University of Chicago Press.
Siles, I., Segura-Castillo, A., Solís, R., & Sancho, M. (2020). Folk theories of algorithmic recommendations on Spotify:
Enacting data assemblages in the global South. Big Data & Society, 7(1), 205395172092337. https://doi.org/10.1177/
2053951720923377
Spotify. (2021). Press center: Spotify Wrapped 2021. For the Record. https://newsroom.spotify.com/media-kit/spotify-
wrapped-2021
Spotify. (2022). Press center: Spotify Wrapped 2022. For the Record. https://newsroom.spotify.com/media-kit/spotify-
wrapped-2022
Spotify. (2023, January 31). Q4 2022 update. https://s29.q4cdn.com/175625835/les/doc_nancials/2022/q4/
Shareholder-Deck-Q4-2022-FINAL.pdf
Stål, O. (2021, November 22). Creating personalized listening experiences. Video. Scale Transform X. https://exchange.
scale.com/public/blogs/inside-the-content-recommendation-engine-at-the-heart-of-spotify
Vacchelli, E. (2018). Embodiment in qualitative research: Collage making with migrant, refugee and asylum seeking
women. Qualitative Research, 18(2), 171–190. https://doi.org/10.1177/1468794117708008
Venema, R., & Lobinger, K. (2017). ‘And somehow it ends up on the internet’. Agency, trust and risks in photo-sharing
among friends and romantic partners. First Monday. https://doi.org/10.5210/fm.v22i7.7860
Webster, J. (2023). The promise of personalisation: Exploring how music streaming platforms are shaping the perfor-
mance of class identities and distinction. New Media & Society, 25(8), 2140–2162. https://doi.org/10.1177/
14614448211027863
Zubo, S. (2019). The age of surveillance capitalism: The ght for a human future at the new frontier of power. Prole Books.
JOURNAL OF GENDER STUDIES 17